Anthropic CCDV-F (Claude Certified Developer - Foundations) Practice Test 53

Question 520

A team maintaining an enterprise MCP client has two different engineering requirements: (1) reject a date or identifier that violates business constraints before a service call, and (2) reduce context bloat caused by a tool returning a thousand irrelevant fields. Which TWO independent changes directly meet them? Choose TWO.

  1. Create a Skill containing the procedure and supporting instructions.
  2. Validate model-supplied tool arguments against schema and domain rules.
  3. Give each tool a timeout and controlled retry or failure result.
  4. Implement a narrowly scoped read-only tool handler.
  5. Return a concise task-relevant projection from the tool handler.

Explanation

Correct Answers

B:

To reject a date or identifier that violates business constraints before a service call as part of a design that must also reduce context bloat caused by a tool returning a thousand irrelevant fields, even syntactically correct generated inputs can violate account scope or allowed values, so execution must validate both levels; this is why the action is one of the two required selections.

E:

To reduce context bloat caused by a tool returning a thousand irrelevant fields as part of a design that must also reject a date or identifier that violates business constraints before a service call, tool output is subsequently consumed as model context; minimizing it reduces both token cost and distractions; this is why the action is one of the two required selections.

Incorrect Answers

A:

This alternative serves the separate goal to make a repeatable multi-step code-review methodology available to an agent because a Skill teaches how to perform a task, whereas a tool is the executable capability that performs an external operation; the required outcomes are instead to reject a date or identifier that violates business constraints before a service call and to reduce context bloat caused by a tool returning a thousand irrelevant fields.

C:

This alternative serves the separate goal to stop a slow external system from trapping an agent loop indefinitely because tool execution is a separate latency source; bounded calls prevent one dependency from exhausting the entire agent budget; the required outcomes are instead to reject a date or identifier that violates business constraints before a service call and to reduce context bloat caused by a tool returning a thousand irrelevant fields.

D:

This alternative serves the separate goal to expose a diagnostic capability without allowing the model to modify records because a restricted read interface supports observation while preventing data mutation, unlike giving a diagnostic agent a general administrative API; the required outcomes are instead to reject a date or identifier that violates business constraints before a service call and to reduce context bloat caused by a tool returning a thousand irrelevant fields.

Question 521

An architecture decision for a model selection trial must support run a straightforward well-specified classification without consuming examples and reduce inconsistent labeling across nuanced examples. Which option gives a complete rather than partial solution?

  1. Perform both steps: (1) Start with a clear instruction and no demonstrations; (2) include several carefully selected and diverse demonstrations.
  2. Perform both steps: (1) Start with a clear instruction and no demonstrations; (2) route simple requests to a cheaper evaluated tier and escalate difficult ones.
  3. Perform both steps: (1) Budget total context for instructions, history, tools, and output; (2) route simple requests to a cheaper evaluated tier and escalate difficult ones.
  4. Perform both steps: (1) Budget total context for instructions, history, tools, and output; (2) include several carefully selected and diverse demonstrations.

Explanation

Correct Answer

A:

The first change enables the team to run a straightforward well-specified classification without consuming examples because zero-shot prompting relies on direct task instructions and can save context when the task requires little format teaching; the second addresses the separate need to reduce inconsistent labeling across nuanced examples because multiple examples communicate decision boundaries and desired formatting, although irrelevant or misleading demonstrations can bias the model.

Incorrect Answers

B:

The first step is appropriate for the requirement to run a straightforward well-specified classification without consuming examples because zero-shot prompting relies on direct task instructions and can save context when the task requires little format teaching; the other step instead aims to minimize average model cost when most requests are easy but some require deeper reasoning because cascading matches compute to task difficulty but requires a reliable escalation rule and measurement of misrouting, which does not meet the second requirement to reduce inconsistent labeling across nuanced examples.

C:

Both steps target different results—know why a long prompt plus a proposed response may exceed the model's accepted context and minimize average model cost when most requests are easy but some require deeper reasoning—because a context window bounds what can be processed in one request; large tool results consume it along with user messages and cascading matches compute to task difficulty but requires a reliable escalation rule and measurement of misrouting; the required capabilities are instead to run a straightforward well-specified classification without consuming examples and to reduce inconsistent labeling across nuanced examples.

D:

The second step meets the requirement to reduce inconsistent labeling across nuanced examples because multiple examples communicate decision boundaries and desired formatting, although irrelevant or misleading demonstrations can bias the model; the first step would only help to know why a long prompt plus a proposed response may exceed the model's accepted context because a context window bounds what can be processed in one request; large tool results consume it along with user messages, leaving the separate requirement to run a straightforward well-specified classification without consuming examples unmet.

Question 522

A change request for a high-volume Messages API gateway combines two needs: design for a documented burst of simultaneous requests without uncontrolled overload, and keep sensitive case material within approved processing and retention boundaries. Which implementation pair addresses both, not just one?

  1. Perform both steps: (1) Cache an eligible unchanged prefix with cache_control; (2) queue or reject excess requests using bounded admission control.
  2. Perform both steps: (1) Specify expected concurrency and enforce admission or queueing limits; (2) queue or reject excess requests using bounded admission control.
  3. Perform both steps: (1) Cache an eligible unchanged prefix with cache_control; (2) document permitted processing regions and data-handling obligations before selecting a deployment.
  4. Perform both steps: (1) Specify expected concurrency and enforce admission or queueing limits; (2) document permitted processing regions and data-handling obligations before selecting a deployment.

Explanation

Correct Answer

D:

The first change enables the team to design for a documented burst of simultaneous requests without uncontrolled overload because capacity planning separates average traffic from peak demand and prevents upstream rate limits from becoming cascading application failures; the second addresses the separate need to keep sensitive case material within approved processing and retention boundaries because legal or organizational constraints must inform provider routing, logging, and retention rather than being added after data already flows.

Incorrect Answers

A:

Both steps target different results—reuse a long stable instruction or document prefix over repeated requests and shield customer traffic when inference capacity is exhausted—because prompt caching reduces repeated input processing when the prefix matches; changes before the cache boundary can prevent a hit and controlled queuing protects system latency and prevents unbounded retries from overwhelming downstream provider quotas; the required capabilities are instead to design for a documented burst of simultaneous requests without uncontrolled overload and to keep sensitive case material within approved processing and retention boundaries.

B:

The first step is appropriate for the requirement to design for a documented burst of simultaneous requests without uncontrolled overload because capacity planning separates average traffic from peak demand and prevents upstream rate limits from becoming cascading application failures; the other step instead aims to shield customer traffic when inference capacity is exhausted because controlled queuing protects system latency and prevents unbounded retries from overwhelming downstream provider quotas, which does not meet the second requirement to keep sensitive case material within approved processing and retention boundaries.

C:

The second step meets the requirement to keep sensitive case material within approved processing and retention boundaries because legal or organizational constraints must inform provider routing, logging, and retention rather than being added after data already flows; the first step would only help to reuse a long stable instruction or document prefix over repeated requests because prompt caching reduces repeated input processing when the prefix matches; changes before the cache boundary can prevent a hit, leaving the separate requirement to design for a documented burst of simultaneous requests without uncontrolled overload unmet.

Question 523

Two acceptance tests block a privileged agent security review: one requires the ability to demonstrate who approved a privileged tool action and when; the other requires the ability to stop frontend users from receiving unrestricted backend tool credentials. Which plan clears both tests?

  1. Perform both steps: (1) Record result metadata in a post-tool hook; (2) use a server-side workload identity or scoped service credential.
  2. Perform both steps: (1) Record authenticated approval events with actor, scope, and request correlation; (2) use a server-side workload identity or scoped service credential.
  3. Perform both steps: (1) Record authenticated approval events with actor, scope, and request correlation; (2) enforce a pre-tool hook that denies the disallowed call.
  4. Perform both steps: (1) Record result metadata in a post-tool hook; (2) enforce a pre-tool hook that denies the disallowed call.

Explanation

Correct Answer

B:

The first change enables the team to demonstrate who approved a privileged tool action and when because an audit trail binds authorization to a specific action, making privilege use reviewable without treating model-generated text as proof of consent; the second addresses the separate need to stop frontend users from receiving unrestricted backend tool credentials because service identities tie privileges to a runtime workload and avoid exposing long-lived secrets to browsers or untrusted users.

Incorrect Answers

A:

The second step meets the requirement to stop frontend users from receiving unrestricted backend tool credentials because service identities tie privileges to a runtime workload and avoid exposing long-lived secrets to browsers or untrusted users; the first step would only help to capture outcomes of agent command execution for security review because post-execution hooks can log and inspect what happened; they are unsuitable as the sole barrier to a side effect that already occurred, leaving the separate requirement to demonstrate who approved a privileged tool action and when unmet.

C:

The first step is appropriate for the requirement to demonstrate who approved a privileged tool action and when because an audit trail binds authorization to a specific action, making privilege use reviewable without treating model-generated text as proof of consent; the other step instead aims to stop a coding agent from running commands that match a prohibited destructive pattern because a deterministic hook checks the proposed command before execution, providing a stronger control than asking the agent to self-police, which does not meet the second requirement to stop frontend users from receiving unrestricted backend tool credentials.

D:

Both steps target different results—capture outcomes of agent command execution for security review and stop a coding agent from running commands that match a prohibited destructive pattern—because post-execution hooks can log and inspect what happened; they are unsuitable as the sole barrier to a side effect that already occurred and a deterministic hook checks the proposed command before execution, providing a stronger control than asking the agent to self-police; the required capabilities are instead to demonstrate who approved a privileged tool action and when and to stop frontend users from receiving unrestricted backend tool credentials.

Question 524

A test for a specialist-agent orchestration service fails because it cannot speed up independent data lookups without causing write-order races. Which change fixes the identified failure?

  1. Run independent branches concurrently and merge their outputs afterward.
  2. Define and validate a typed schema at the handoff.
  3. Retrieve only task-relevant memories into the active context.
  4. Run side-effect-free tool calls concurrently when their inputs are independent.

Explanation

Correct Answer

D:

To speed up independent data lookups without causing write-order races in this specialist-agent orchestration service, read-only independent calls can safely overlap; mutating operations need explicit ordering or concurrency control; the proposed implementation directly supplies that capability.

Incorrect Answers

A:

The proposed alternative helps to reduce elapsed time for independent extraction tasks that do not share mutable state because parallel orchestration shortens the critical path when calls have no dependencies, while the merger checks completeness before proceeding; that mechanism does not meet the different requirement to speed up independent data lookups without causing write-order races.

B:

The proposed alternative helps to pass reliable machine-readable outputs between independently developed agent stages because typed outputs make downstream contracts inspectable and reduce ambiguity caused by free-form prose passed as control data; that mechanism does not meet the different requirement to speed up independent data lookups without causing write-order races.

C:

The proposed alternative helps to stop old agent histories from burying the information needed for the current step because selective retrieval improves signal-to-noise and keeps token use bounded while retaining access to older information when required; that mechanism does not meet the different requirement to speed up independent data lookups without causing write-order races.

Question 525

A team maintaining a production model-assisted feature must meet 2 separate acceptance criteria: (1) detect rising inference errors after a feature reaches production; (2) process a large queue of independent, noninteractive requests at a lower-cost asynchronous pace. Which two implementation actions directly satisfy the stated requirements? Choose TWO.

  1. Submit independent requests through the Message Batches API.
  2. Track latency, failure rates, token use, and output-quality signals over time.
  3. Store prompt versions with release identifiers and rollback history.
  4. Handle stop_reason max_tokens as truncation and continue or revise the limit.
  5. Version prompt files alongside relevant application code.

Explanation

Correct Answers

A:

The test to process a large queue of independent, noninteractive requests at a lower-cost asynchronous pace is met because batch processing is suited to deferred work with eventual result collection and is not a substitute for an interactive low-latency chat endpoint; the remaining selection covers the separate need to detect rising inference errors after a feature reaches production.

B:

The test to detect rising inference errors after a feature reaches production is met because production health spans infrastructure and model behavior, so a healthy HTTP server alone does not prove the AI feature is functioning; the remaining selection covers the separate need to process a large queue of independent, noninteractive requests at a lower-cost asynchronous pace.

Incorrect Answers

C:

The action is useful for the separate capability to determine whether a quality regression came from a recent instruction edit because a prompt is a behavior-changing artifact; versioning makes comparisons and safe reversions possible during incident response; however, the required test outcomes are to process a large queue of independent, noninteractive requests at a lower-cost asynchronous pace and to detect rising inference errors after a feature reaches production.

D:

The action is useful for the separate capability to recover when a response ends because its generation budget was reached because a truncated completion may contain unfinished content, so treating it as a validated final answer risks losing required fields; however, the required test outcomes are to process a large queue of independent, noninteractive requests at a lower-cost asynchronous pace and to detect rising inference errors after a feature reaches production.

E:

The action is useful for the separate capability to trace exactly which prompt template was deployed with a code change because prompts change software behavior and should be reviewed and diffed like code, rather than edited anonymously in production; however, the required test outcomes are to process a large queue of independent, noninteractive requests at a lower-cost asynchronous pace and to detect rising inference errors after a feature reaches production.

Question 526

The technical owner of a model-family migration plan has to solve demanding multistep reasoning tasks where quality matters more than lowest unit price. A separate requirement is to choose between two supported model releases using evidence rather than reputation. Which two-part remediation covers both?

  1. Perform both steps: (1) Benchmark an Opus-family model on the hardest cases; (2) design an approved fallback model path with compatibility checks.
  2. Perform both steps: (1) Benchmark an Opus-family model on the hardest cases; (2) compare both candidates on identical representative eval cases.
  3. Perform both steps: (1) Configure supported adaptive thinking behavior rather than forcing a fixed pattern; (2) design an approved fallback model path with compatibility checks.
  4. Perform both steps: (1) Configure supported adaptive thinking behavior rather than forcing a fixed pattern; (2) compare both candidates on identical representative eval cases.

Explanation

Correct Answer

B:

The first change enables the team to solve demanding multistep reasoning tasks where quality matters more than lowest unit price because higher-capability tiers may improve difficult-task performance, but their cost and latency still need objective justification; the second addresses the separate need to choose between two supported model releases using evidence rather than reputation because holding inputs and scoring constant reveals quality and latency changes that generic model descriptions cannot settle.

Incorrect Answers

A:

The first step is appropriate for the requirement to solve demanding multistep reasoning tasks where quality matters more than lowest unit price because higher-capability tiers may improve difficult-task performance, but their cost and latency still need objective justification; the other step instead aims to preserve critical service during a temporary primary-model problem because fallbacks improve resilience only when downstream schemas, cost limits, and policy controls work with the alternate model, which does not meet the second requirement to choose between two supported model releases using evidence rather than reputation.

C:

Both steps target different results—allow supported models to vary internal reasoning effort according to task difficulty and preserve critical service during a temporary primary-model problem—because adaptive reasoning responds to problem complexity, but supported modes differ by model and release and should not be assumed universal and fallbacks improve resilience only when downstream schemas, cost limits, and policy controls work with the alternate model; the required capabilities are instead to solve demanding multistep reasoning tasks where quality matters more than lowest unit price and to choose between two supported model releases using evidence rather than reputation.

D:

The second step meets the requirement to choose between two supported model releases using evidence rather than reputation because holding inputs and scoring constant reveals quality and latency changes that generic model descriptions cannot settle; the first step would only help to allow supported models to vary internal reasoning effort according to task difficulty because adaptive reasoning responds to problem complexity, but supported modes differ by model and release and should not be assumed universal, leaving the separate requirement to solve demanding multistep reasoning tasks where quality matters more than lowest unit price unmet.

Question 527

An end-to-end test for a function-calling service checks whether it can expose a diagnostic capability without allowing the model to modify records and reduce context bloat caused by a tool returning a thousand irrelevant fields. Which plan should be approved to satisfy both checks?

  1. Perform both steps: (1) Implement a narrowly scoped read-only tool handler; (2) define a descriptive tool name and validated input schema.
  2. Perform both steps: (1) Describe each tool's purpose, required inputs, and limitations clearly; (2) return a concise task-relevant projection from the tool handler.
  3. Perform both steps: (1) Describe each tool's purpose, required inputs, and limitations clearly; (2) define a descriptive tool name and validated input schema.
  4. Perform both steps: (1) Implement a narrowly scoped read-only tool handler; (2) return a concise task-relevant projection from the tool handler.

Explanation

Correct Answer

D:

The first change enables the team to expose a diagnostic capability without allowing the model to modify records because a restricted read interface supports observation while preventing data mutation, unlike giving a diagnostic agent a general administrative API; the second addresses the separate need to reduce context bloat caused by a tool returning a thousand irrelevant fields because tool output is subsequently consumed as model context; minimizing it reduces both token cost and distractions.

Incorrect Answers

A:

The first step is appropriate for the requirement to expose a diagnostic capability without allowing the model to modify records because a restricted read interface supports observation while preventing data mutation, unlike giving a diagnostic agent a general administrative API; the other step instead aims to expose a business lookup with predictable typed arguments because well-described schemas help Claude select a capability and emit usable arguments; the backend still validates authorization and meaning, which does not meet the second requirement to reduce context bloat caused by a tool returning a thousand irrelevant fields.

B:

The second step meets the requirement to reduce context bloat caused by a tool returning a thousand irrelevant fields because tool output is subsequently consumed as model context; minimizing it reduces both token cost and distractions; the first step would only help to prevent two similarly named lookup tools from being confused because tool descriptions influence the model's selection, so overlapping or vague names make dispatch less reliable, leaving the separate requirement to expose a diagnostic capability without allowing the model to modify records unmet.

C:

Both steps target different results—prevent two similarly named lookup tools from being confused and expose a business lookup with predictable typed arguments—because tool descriptions influence the model's selection, so overlapping or vague names make dispatch less reliable and well-described schemas help Claude select a capability and emit usable arguments; the backend still validates authorization and meaning; the required capabilities are instead to expose a diagnostic capability without allowing the model to modify records and to reduce context bloat caused by a tool returning a thousand irrelevant fields.

Question 528

A browser-based Claude product must keep its provider API key out of client code, and long chat transcripts are pushing the important instructions out of context. Which TWO targeted implementation changes address those distinct risks? Choose TWO.

  1. Configure supported structured outputs with a JSON Schema.
  2. Route Claude requests through a trusted backend service.
  3. Apply a context retention policy and summarize earlier turns where suitable.
  4. Enforce per-tenant token and request budgets in the backend.
  5. Enable required plugin or connector capabilities using least-privilege configuration.

Explanation

Correct Answers

B:

To keep a browser application from exposing the secret used for model calls as part of a design that must also stay within input limits while keeping the most important conversation constraints, a backend controls credentials and access policy while the UI sends user input through an authenticated application endpoint; this is why the action is one of the two required selections.

C:

To stay within input limits while keeping the most important conversation constraints as part of a design that must also keep a browser application from exposing the secret used for model calls, selective history management preserves task-critical state; dropping random leading messages can erase consent or requirements; this is why the action is one of the two required selections.

Incorrect Answers

A:

This alternative serves the separate goal to need schema-conformant JSON for a downstream parser at generation time because constrained JSON generation targets machine-parseable fields and types; free-form prompting alone does not establish the same guarantee; the required outcomes are instead to keep a browser application from exposing the secret used for model calls and to stay within input limits while keeping the most important conversation constraints.

D:

This alternative serves the separate goal to keep one customer from consuming an entire shared inference budget because application quotas isolate noisy tenants and control spending; provider-wide rate limits alone do not guarantee fairness between customers; the required outcomes are instead to keep a browser application from exposing the secret used for model calls and to stay within input limits while keeping the most important conversation constraints.

E:

This alternative serves the separate goal to expose external business tools only to workflows that actually need them because tools expand the application's action surface; indiscriminate activation increases risk and can confuse model selection among similar functions; the required outcomes are instead to keep a browser application from exposing the secret used for model calls and to stay within input limits while keeping the most important conversation constraints.

Question 529

A planned upgrade to a structured output service has two requirements that must both pass: avoid a prohibited output element such as confidential identifiers; teach a classifier how to label borderline requests consistently. Identify the two direct implementations. Choose TWO.

  1. State explicit priority and conflict resolution for the task constraints.
  2. Specify the forbidden content alongside an enforceable validation check.
  3. Supply representative labeled examples including edge conditions.
  4. Name each mandatory field or section and provide formatting constraints.
  5. Summarize prior decisions and preserve unresolved constraints.

Explanation

Correct Answers

B:

To avoid a prohibited output element such as confidential identifiers as part of a design that must also teach a classifier how to label borderline requests consistently, negative prompt guidance signals the intended boundary, but a deterministic validator is needed where violating it has consequences; this is why the action is one of the two required selections.

C:

To teach a classifier how to label borderline requests consistently as part of a design that must also avoid a prohibited output element such as confidential identifiers, diverse demonstrations make the intended decision boundary clearer than an isolated easy example, provided labels are accurate; this is why the action is one of the two required selections.

Incorrect Answers

A:

This alternative serves the separate goal to prioritize mandatory legal requirements over optional stylistic preferences because without a priority rule, competing directives can create ambiguous outputs; hard requirements should guide how optional preferences are applied; the required outcomes are instead to avoid a prohibited output element such as confidential identifiers and to teach a classifier how to label borderline requests consistently.

D:

This alternative serves the separate goal to require that a generated report have a fixed set of sections because concrete desired structure is easier to verify than a vague instruction to be professional or comprehensive; the required outcomes are instead to avoid a prohibited output element such as confidential identifiers and to teach a classifier how to label borderline requests consistently.

E:

This alternative serves the separate goal to maintain the current task after older dialogue exceeds the active token budget because a faithful rolling summary carries forward objectives and commitments while avoiding a full-history replay on each call; the required outcomes are instead to avoid a prohibited output element such as confidential identifiers and to teach a classifier how to label borderline requests consistently.

Question 530

An end-to-end test for a token-consumption optimization sprint checks whether it can build bidirectional persistent event communication where the chosen platform supports it and process partial model output as it is delivered over a streaming connection. Which plan should be approved to satisfy both checks?

  1. Perform both steps: (1) Provide one accurate input-output demonstration; (2) model the response as conditioned token generation.
  2. Perform both steps: (1) Provide one accurate input-output demonstration; (2) iterate the sdk or http event stream until its terminal event.
  3. Perform both steps: (1) Use a WebSocket protocol only when the target endpoint actually offers one; (2) iterate the sdk or http event stream until its terminal event.
  4. Perform both steps: (1) Use a WebSocket protocol only when the target endpoint actually offers one; (2) model the response as conditioned token generation.

Explanation

Correct Answer

C:

The first change enables the team to build bidirectional persistent event communication where the chosen platform supports it because webSockets can carry two-way events over a long-lived connection; they are not a drop-in replacement for every HTTP or SSE endpoint; the second addresses the separate need to process partial model output as it is delivered over a streaming connection because streaming clients must handle both content deltas and final status rather than assuming the first chunk is the complete response.

Incorrect Answers

A:

Both steps target different results—teach a distinctive output format with one representative example and explain why language-model output is generated step by step rather than fetched from a fixed response table—because a single example illustrates structure and conventions, but it may not reveal how to handle difficult edge cases and autoregressive decoding selects subsequent tokens based on supplied context and prior output, which creates flexible but non-guaranteed wording; the required capabilities are instead to build bidirectional persistent event communication where the chosen platform supports it and to process partial model output as it is delivered over a streaming connection.

B:

The second step meets the requirement to process partial model output as it is delivered over a streaming connection because streaming clients must handle both content deltas and final status rather than assuming the first chunk is the complete response; the first step would only help to teach a distinctive output format with one representative example because a single example illustrates structure and conventions, but it may not reveal how to handle difficult edge cases, leaving the separate requirement to build bidirectional persistent event communication where the chosen platform supports it unmet.

D:

The first step is appropriate for the requirement to build bidirectional persistent event communication where the chosen platform supports it because webSockets can carry two-way events over a long-lived connection; they are not a drop-in replacement for every HTTP or SSE endpoint; the other step instead aims to explain why language-model output is generated step by step rather than fetched from a fixed response table because autoregressive decoding selects subsequent tokens based on supplied context and prior output, which creates flexible but non-guaranteed wording, which does not meet the second requirement to process partial model output as it is delivered over a streaming connection.

Claude Certified Developer - Foundations (CCDV-F) Practice Tests